3 ms·
iRobot's failure is that they made a bet to use CV instead of Lidar for their mapping robots for a long time until it was too late. That made their affordable,
by sudosysgen 10mo ago
iRobot's failure is that they made a bet to use CV instead of Lidar for their mapping robots for a long time until it was too late. That made their affordable, non-mapping robots far far worse than only slightly higher priced lidar robots, while their mapping robots were too expensive for mass appeal and were still worse at navigation than up-market lidar based robots. Ultimately they were simply outcompeted.
- CrossVR 10mo agoReminds me of a certain self-driving car company.
- dzhiurgis 10mo agoIs it one with P/E > 300?
- Animats 10mo agoNeato, which had a robot vacuum with LIDAR, shut down in 2023. That's not the key problem. Binocular vision ought to be good enough for a vacuum. It's short range compared to the inter-camera distance. Vehicle object ranging at distance is much tougher and can be fooled.
- sudosysgen 10mo ago> Binocular vision ought to be good enough for a vacuum. It could be, but it just is not. VSLAM robots were practically significant worse. There are a lot of limitations to multi-ocular vision for a robot vacuum, for example the relatively featureless walls and few features across the horizontal binocular axis. Neato was never as big as iRobot. They didn't fail from commanding heights, they never were that successful to begin with, for entirely different reasons. If they had managed to get to iRobot's level of ubiquity and distribution they would have had a much better shot of still being around nowadays.
- Animats 10mo ago> relatively featureless walls Right. The cheap solution to that is projecting a pattern of IR dots on the walls to give them some features. One version of Microsoft's Kinect did that.
- aschobel 10mo agoDidn't Matic solve this (non-Lidar robot vacuum)? People seem to rave about them.
- kobieps 10mo agoYeah matic is vision only